"We don't have time for customer interviews" - what 9 conversations actually changed

"We don't have time for customer interviews."

I hear that sentence more than any other objection in product, and it is almost never about the calendar. It is a belief: that building is progress and talking to customers is overhead you pay for later, when things calm down.

Things never calm down. So here is a receipt instead of an argument - what nine conversations cost, what they found, and the two numbers that moved afterwards.

Nine interviews, about five hours

At UX Pilot, a bootstrapped AI design tool with more than a million users, we ran nine in-depth interviews with people on the Pro plan. All senior designers, spread across different industries.

The goal was deliberately not "do you like the product". It was to understand the ideal customer: their characteristics, their needs, and the specific use cases they kept coming back for.

Total cost: five hours of calls, spread over a couple of weeks. No engineering time. No release. No designer pulled off a project.

What nine conversations found

All nine described the same primary use case: accelerating solution exploration during the design process. Not one of them led with anything else.

One put it precisely enough that I wrote it down verbatim:

Another had replaced their entire whiteboard, then sketches, then wireframes ideation phase with it.

When nine people who have never met independently describe the same job, that is not a preference you have collected. It is the spine of the product. And the corollary matters just as much: everything else we had been treating as core value turned out to be secondary.

The change we made was small. That is the point.

We did not build a new feature. We made generating design variations and importing components easier to discover and easier to use.

Two numbers moved in the weeks after:

Here is what I want you to notice. That change would never have won an argument in a prioritisation meeting. It was not a feature. On a backlog it reads like a polish ticket - "improve discoverability of variations" - sitting below four things with revenue attached to them.

It only became obvious once nine people had told us what they were actually trying to do, and we could see the gap between that job and the path the product offered them.

The excuse costs more than the interviews

The usual case for research is defensive: it stops you building the wrong thing. True, and I have written about the arithmetic of five conversations versus a wasted quarter before.

But the defensive case undersells it. The bigger loss is the improvement you never find. High-leverage changes are often small, unglamorous, and invisible from inside the building, because they live in the gap between what your product does and what your customer is trying to get done. No amount of backlog grooming surfaces them. Only someone describing their actual work does.

So when a team says there is no time for interviews, they are not choosing speed over rigour. They are choosing to keep guessing which of their ideas is the good one, and to find out by shipping.

"No time" almost always means "no habit"

Three tells, and you will recognise all of them:

• Research is treated as a project. It needs a kickoff, a plan, and a stakeholder, so it needs a quarter that has room in it. There is never a quarter with room in it.
• Nobody owns recruiting. Everyone agrees interviews are valuable and nobody has a list of people to call, so every round starts from zero.
• It is not in the calendar. Anything that is not a standing slot competes with delivery every week, and loses every week.

The fix is unglamorous: two conversations a week, booked as a recurring commitment, forever. You do not need a research function. You need a calendar invite and someone whose job it is to fill it.

How to run this exact study yourself

Pick one segment. Your best-fitting paying users, not a cross-section of everyone. Nine designers on the same plan told us more than ninety mixed users would have, because their answers did not average out into mush.

Recruit eight to ten of them. Screen for the problem, not for enthusiasm. People who love you are pleasant company and weak evidence.

Ask about the last time, never the next time. "Walk me through the last time you did this. What did you try first? Why did that not work?" The moment you ask what someone wants, you are back in the Product Death Cycle, collecting solutions from people who were never responsible for designing them.

Look for the repeated job, not the popular request. Requests scatter. Jobs converge. The thing several people describe in different words is the one worth acting on.

Instrument the change before you ship it. Decide which number should move, and by roughly how much, while you still have the humility to be wrong. That is the difference between learning and storytelling, and it is the whole basis of building on evidence rather than a hunch.

Installing that loop as a weekly habit, rather than a one-off study, is the first thing we do in a Product Discovery Workshop. It is also the part teams are most surprised to find they can sustain.

The reply I would give

Next time someone says there is no time for customer interviews, do not argue about principles. Ask what the last five hours of customer conversation produced.

If the answer is nothing, it is not because five hours was too expensive. It is because nobody spent them.